broadbrained Space, news, venture, robots, technology
In order to prevent collisions in multi-robot systems, constant broadcasting of each robot trajectory plans is often employed. However, this requires a lot of data transfer and is not necessary when the trajectories are clearly not intersecting.
A recent paper suggests an efficient communication policy and trajectory planning method for micro-aerial vehicles. Multi-Agent Reinforcement Learning is used to determine when and with whom it is most useful to communicate. Using requested and estimated trajectories of other robots, the safest trajectory is decided.
Drone. Image credit: Pxhere, CC0 Public Domain
The method was verified in a simulation using four drones in the scenarios of different complexity. It is observed that the policy triggers communication either at the beginning of the motion or when the drones are in a collision course. The policy reduced the number of communications requests approximately 77% and achieved zero collisions at the same time.
Decentralized multi-robot systems typically perform coordinated motion planning by constantly broadcasting their intentions as a means to cope with the lack of a central system coordinating the efforts of all robots. Especially in complex dynamic environments, the coordination boost allowed by communication is critical to avoid collisions between cooperating robots. However, the risk of collision between a pair of robots fluctuates through their motion and communication is not always needed. Additionally, constant communication makes much of the still valuable information shared in previous time steps redundant. This paper presents an efficient communication method that solves the problem of “when” and with “whom” to communicate in multi-robot collision avoidance scenarios. In this approach, every robot learns to reason about other robots’ states and considers the risk of future collisions before asking for the trajectory plans of other robots. We evaluate and verify the proposed communication strategy in simulation with four quadrotors and compare it with three baseline strategies: non-communicating, broadcasting and a distance-based method broadcasting information with quadrotors within a predefined distance.
Link: https://arxiv.org/abs/2009.12106
